Control of energy costs is a high priority with businesses and governments. The assessment of a baseline energy usage profile for a building as related to mechanical systems consumption due to heating and cooling loads is fairly well understood according to building science principles. The baseline energy costs may be inaccurate due to unverified and unreported occupant and systems behavior which is not directly discoverable. Furthermore, there are a large number of variables involved in the modeling of energy consumption such as occupant behavioral factors and unknown equipment efficiencies that feed inaccuracy of the results and lead to poor decision making. Calibration of the baseline energy usage profile to historical energy usage is possible; however, calibration methods are not an exact science due to the large number of variables involved.
U.S. Pat. No. 6,134,511 to Subbarao discloses a method and apparatus for improving building energy simulations where the calibration of building energy simulations with performance data is accomplished by introducing corrective heat flows. Subbarao utilizes the energy simulator DOE-2 which requires complex evaluation of a large number of inputs and outputs and is not suitable for providing rapid feedback.
U.S. Pat. No. 6,968,295 to Carr discloses a method of and system for auditing the energy-usage by a facility, where the facility includes an energy-using system having an operational parameter with a value. Carr does not disclose a calibration process for energy-usage.
U.S. Pat. No. 7,881,889 to Barclay et al. discloses a computer implemented method to facilitate determining energy cost savings in an energy consuming facility using an artificial intelligence model. The drawback of Barclay et al. is that the disclosed method requires a wide variety of training data sets to predict energy savings accurately.
U.S. Patent Publication No. 2011/0153103 to Brown et al. discloses a system and method for predictive modeling of building energy consumption providing predicted building energy load values determined by smoothing of historical building energy load values for a building. Brown et al. requires complex optimization training to optimize prediction of building energy load values by “cross-validation error minimization.”
U.S. Patent Publication No. 2011/0246381 to Fitch et al. discloses a method of modeling energy usage and cost impacts for a building and comparing a theoretical data set to an actual building performance to determine a margin of error. Fitch et al. requires complex evaluation of a large number of inputs and outputs.
U.S. Patent Publication No. 2011/0251933 to Egnor et al. discloses a system and method for modeling a building's energy usage over time based on historic data. Egnor et al. uses a regression analysis which requires extensive data sets for predicting the energy usage of the building.
U.S. Patent Publication No. 2012/0084063 to Drees et al. discloses a system for detecting changes in energy usage in a building. A baseline energy usage model is determined from a least squares regression analysis.